azure-ai-openai-dotnet
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a technical reference for the official Microsoft Azure OpenAI .NET SDK. It follows standard documentation patterns and does not contain any malicious instructions, obfuscation, or unauthorized data access.- [EXTERNAL_DOWNLOADS]: All package installation instructions (e.g.,
dotnet add package Azure.AI.OpenAI) refer to official, well-known libraries from Microsoft and OpenAI hosted on the standard NuGet registry.- [PROMPT_INJECTION]: The skill identifies surfaces for processing untrusted data through LLM interactions, which is inherent to its purpose. - Ingestion points: Untrusted data enters the context via
UserChatMessageandAzureSearchChatDataSource(RAG) as shown inSKILL.md. - Boundary markers: The code snippets utilize standard SDK role-based message types (
SystemChatMessage,UserChatMessage,AssistantChatMessage) to provide structural separation. - Capability inventory: The documented capabilities include network access to Azure OpenAI endpoints, local file system writes for audio output (
output.mp3), and tool execution for function calling. - Sanitization: The skill explicitly recommends best practices such as validating tool call arguments before execution and using structured JSON schema outputs to ensure data integrity.
Audit Metadata